Astrocytomas, including low grade astrocytoma (grade II), anaplastic (high grade) astrocytoma (grade III) and glioblastoma (grade IV), are the most common human brain tumor. Primarily because of their diffuse nature, there is no effective treatment for these cancers, and little is known about the mechanism(s) by which they develop. For example, we do not understand the nature of susceptible target cell(s) and microenvironment influences, the cause and effect relationships between genetic lesions and cell/tissue responses, nor the molecular and cellular basis for progression from low to high-grade malignancy. Because even normal developmental and homeostatic mechanisms are difficult to discern within the intricate environment of the mammalian brain, experimental animal models are required to understand the complexities of astrocytic cancers and to develop reasonable therapies for its treatment. Recent advances in the areas of brain cell lineages, astrocyte biology, human astrocytoma genetics, and the development of appropriate preclinical mouse models have provided a foundation for addressing these problems. Furthermore, such studies are facilitated by emerging technological advances in mouse modeling, genomics and imaging methodologies. The current proposal aims to launch a comprehensive assessment of astrocytic tumorigenesis by drawing on these diverse areas. Here we will probe the biological consequences of genetic alterations introduced into specific cellular targets and within specific microenvironments of the brain. Biological analysis will include molecular and cell biological approaches coupled with cellular, tissue and live animal imaging. The PI (Van Dyke) and co-PI (Gutmann) have already developed important GEM astrocytoma models utilizing distinct strategies. The former model develops GEM grade III astrocytoma with complete penetrance after inactivation of the pRb pathway in the astrocytic lineage, while the latter develops high grade astrocytoma and glioblastoma upon Ras activation within the same lineage. While we continue to probe basic genetic and biological mechanisms of astrocytic cancers we will also utilize genomic and imaging tools to monitor the extent of similarities between the mouse and human diseases. In addition we will work to establish databases and other sharing mechanisms for the data generated by us and others. Collaborators in these efforts include Drs. Ellisman (cellular and tissue imaging;database development;UCSD), Johnson (small animal imaging;Duke University);Bullitt (imaging of vessel properties;UNC);Lin (MRI analysis of tumor properties;UNC);Baker (expression analysis;U. Florida) and Pevney (developmental neurobiology;UNC).